1. Field of the Invention
The present invention relates to data communications. In particular, the present invention relates to generating jitter test patterns on a high performance serial bus system.
2. The Prior Art
Background
The effect of jitter on the transmission of electrical signals is known in the art. Jitter is generally defined as the deviation of a signal edge from its ideal position. Jitter is generally introduced by the electrical properties of the transmitting device and the transmission medium. For example, long runs of cable add jitter as they affect the rise and fall times and amplitude of a signal due to the time constants created by the additional capacitance inherent in a long run of cable.
Furthermore, jitter can be dependent upon the type of signal being transported. For example, a long string of zeros followed by a single one can cause problems because charging a long run of cable is difficult in such a short amount of time. Likewise, a long string of ones followed by a single zero can also cause problems because of the difficulty in discharging a long run of cable in such a short period of time. The longer run length produces a higher amplitude which takes more time to overcome when changing bit values and therefore produces a time difference compared to the run length of a 1 bit sequence. When different run lengths are mixed in the same transmission the different bit sequences (symbols) therefore interfere with each other. This effect is known as intersymbol interference (ISI). ISI is expected whenever any bit sequence has frequency components that are propagated at different rates by the transmission media.
Thus, for any communications system to operate effectively, the transmitter and transmission medium must limit the jitter introduced into the signal, and receiver must be able to tolerate any permitted jitter introduced into the signal. The presence of jitter complicates the design of receiver circuits which are required to use clock recovery techniques, for example Phase Locked Loops (PLL) or Delay Locked Loops (DLL). Such circuits are often found to be particularly sensitive to the frequency of the jitter. The frequency of the jitter is a property of the frequency of the repeat patterns in the data being transmitted combined with the effects of ISI and other sources of jitter.
Attempts have been made in the prior art to characterize jitter and generate jitter test patterns which correspond the worst-case scenario that a receiver might face. Communications equipment can then be subjected to the jitter test patterns to examine the communications equipment""s susceptibility to jitter effects. The results may then be plotted for analysis.
FIG. 1 shows a prior art jitter diagram known in the art as an eye diagram. The eye diagram of FIG. 1 is typically displayed on an oscilloscope with a storage facility to store and display a large number of signals simultaneously or a communications analyzer, with the vertical axis representing voltage, and the horizontal axis representing time.
The eye diagram of FIG. 1 includes a low potential level 100 (xe2x80x9clowxe2x80x9d) and a high potential level 102 (xe2x80x9chighxe2x80x9d) which correspond to the absolute value of a logical low and high state, respectively. A threshold 104 is defined on the eye diagram of FIG. 1. Threshold 104 corresponds to the voltage level above which a receiving device will sense a logical high, and below which a device will sense a logical low.
FIG. 1 further includes a crossing 0 (106) and a crossing 1 (108). Crossings 0 and 1 define the left and rightmost boundaries of bit window 110, respectively. As is appreciated by those of ordinary skill in the art, it is desirable that all transitions should occur outside of the bit window 110. In actual practice, the effect of jitter is that the transitions may occur at a variety of times, as shown by transitions 1 . . . N (107). However, as long as the transitions do not encroach on the bit window 110, the receiving device will be able to accurately decode the information.
As is known by those of ordinary skill in the art, jitter is indicated by distributed transitions (crossings) of the threshold as the data toggles between logic states. Using equipment standard in the art such as a time interval analyzer (TIA), histograms of transition regions can be taken at the threshold level. The width of the histograms can then be estimated using methods and algorithms standard in the art, including standard deviation, etc.
One area that is impacted by jitter effects is high performance serial buses. One such bus protocol is the IEEE 1394-1995 standard. This standard revolutionized the consumer electronics industry by providing a serial bus management system that featured high speeds and the ability to xe2x80x9chotxe2x80x9d connect equipment to the bus; that is, the ability to connect equipment without first turning off the existing connected equipment. Since its adoption, the IEEE 1394-1995 standard has begun to see acceptance in the marketplace with many major electronics and computer manufacturers providing IEEE 1394-1995 connections on equipment that they sell.
The IEEE 1394-1995 standard was not greatly impacted by jitter effects, however, because of the relatively short cable lengths (about 3 meters, maximum) utilized by the standard and the transmission of a clock reference by use of Data-Strobe encoding.
However, as technologies improved, the need to update the IEEE 1394-1995 standard became apparent. A new standard is being proposed at the time of the filing of this application, herein referred to as the P1394b standard. Improvements such as higher speeds and longer connection paths will be provided. It is contemplated at the time of this filing that cable lengths exceeding 100 meters may be possible using the P1394b standard. In addition, the use of Data Strobe encoding is impractical at these higher frequencies and cable lengths, and so P1394b requires receivers to use clock recovery techniques by analyzing the timing of the data edges of the incoming signal. Accordingly, it is possible that systems operating under the P1394b standard may be susceptible to jitter effects.
In the discussion that follows, it will be necessary to distinguish between the various standards that are being proposed as of the date of this application. Thus, the term xe2x80x9cLegacyxe2x80x9d will be used herein to refer to the IEEE 1394-1995 standard and all supplements thereof prior to the P1394b standard. Thus, for example, a Legacy node refers to a node compatible with the IEEE 1394-1995 standard and all supplements thereof up to, but not including, the P1394b standard.
As mentioned above, the Legacy standard was not greatly impacted by jitter effects. This is evidenced by a very simple jitter specification and lack of jitter test patterns (for example, to take into account the effect of ISI) in the Legacy standard as adopted. However, because of the performance increases in both speed and run length and the need to use clock recovery techniques, implementations of the P1394b standard may be susceptible to jitter effects.
Hence, there is a need for a system for measuring and determining the effects of jitter upon devices compliant with the P1394b standard. Furthermore, there is a need for a method for generating jitter test patterns within the P1394b environment.
Other industries have also attempted to measure and characterize jitter effects. One such industry is the fiber optics industry, and in particular, the Fibre Channel (FC) link used within the fiber optics industry. The various manufacturers of Fibre Channel-compliant hardware have produced a document entitled xe2x80x9cMethodologies for Jitter Specification, Draft Proposed Technical Report, Secretariat National Committee for Information Technology Standardization (NCITS)xe2x80x9d, hereinafter xe2x80x9cMJSxe2x80x9d. MJS is an ANSI technical report on the definitions, measurement requirements, and allowed values of jitter on a 1.0625 GBaud Fibre Channel link.
The results of the MJS provide an excellent framework to begin developing jitter test patterns. For example, for the Fibre Channel jitter tolerance test contained in the MJS, the following assumptions were made: 1) the average FC traffic transition density is approximately 50%; 2) the CDR time constant is inversely proportional to transition density; 3) to obtain at least 95% settling a pattern duration needs to be greater than 3 time constants 4) The PLL""s minimum bandwidth for FC transceivers is 637 kHz.
However, the test patterns recommended in the MJS are specific to the Fibre Channel encoding and frame format, as can be seen from the FC-specific assumptions laid out above. As such, these assumptions were made in the MJS that are specific to the hardware and software used in the Fibre Channel media. Thus, while the goals of the MJS are desirable, the solutions provided in the MJS cannot be implemented in the P1394b standard because of the differences in encoding and frame format between the two industries.
The invention satisfies the above needs. The present invention relates to data communications. In particular, the present invention relates to generating jitter test patterns on a high performance serial bus system.
Jitter pattern tests and protocols are disclosed herein which may be generated by test software, but run on a normal P1394b node with no special hardware modification or test modes.
A first aspect of the present invention provides a method for generating random jitter test patterns comprising generating a sequence of maximum size asynchronous packets according to the P1394b standard and transmitting the sequence to the device under test. Other embodiments of the first aspect include a test sequence which comprises CURRENT_ASYNC requests, ignoring fairness; and where the sequence comprises a packet payload containing all zeros.
A second aspect of the present invention provides for a method for generating jitter test patterns comprising disabling the transmitter data scrambler of the second device; clearing the port_error register of the device under test; and sending a test pattern to said device under test.
Further embodiments of the second aspect of the present invention provide that the error rate of a jitter test may be computed by reading the port_error register. Further embodiments of the test pattern comprise a sequence of at least 42 null packets followed by a packet containing at least 50 bytes of consecutive D21.5 symbols having an 8 bit data value AD16, and where the sequence is generated by repeatedly issuing CURRENT_ASYNC requests and providing a zero length PHY packet.
A third aspect of the present invention provides for a method for generating supply noise test patterns comprising transmitting a test pattern to the DUT comprising a maximum length asynchronous packet containing alternate 0016 and FF16 bytes.
A fourth aspect of the present invention provides for a method for generating supply noise test patterns comprising transmitting data across a system bus electrically proximate to the device under test and the second device, said data alternating between all ones and all zeros on successive clocks.